Pelvic Obliquity Correction in Distraction-Based Growth Friendly Implants.

DESIGN

Multicenter retrospective review.

OBJECTIVE

To evaluate radiographic outcomes and complication rates of patients treated with distraction based implants and pelvic fixation with either screws (sacral-alar-iliac [SAI] screws or iliac screws) or hooks (S hook iliac fixation).

SUMMARY OF BACKGROUND DATA

Multiple options exist for pelvic fixation in distraction-based growing rod systems; however, limited comparative data are available.

METHODS

Early-onset scoliosis (EOS) patients of all diagnoses with distraction-based implants that had pelvic fixation from 2000 to 2013 were reviewed from two EOS multicenter databases. Patients were divided into two groups by type of pelvic fixation: (1) screw group (SAI screws or iliac screws) or (2) S hooks. Exclusion criteria were as follows: index instrumentation ≥10 years old and follow up <2 years. A total of 153 patients met the inclusion criteria. Mean age at index surgery was 6.1 years (range 1.0-9.9 years) and mean follow-up was 4.9 years.

RESULTS

Pelvic fixation in the 153 patients was as follows: screw group = 42 and S hook group = 111. When comparing patients with >20° of initial pelvic obliquity, the screw group had significantly more correction; mean 26° ± 13° for the screw group versus mean 17° ± 7° in the S hook group (p = .039). There was no significant difference in change in T1-S1 length (40 vs. 39 mm, p = .89) or correction of Cobb angle (30° vs. 24°, p = .24). The total complication rate for the screw group was 14% (6/42) versus 25% (28/111) in the S hook group, though this did not achieve significance (p = .25). The most common complications were device migration (13), implant failure (8), and implant prominence (4) for S hooks and implant failure (3), implant prominence (2), and device migration (1) for the screw group.

CONCLUSION

In distraction-based growth-friendly constructs, pelvic fixation with screws achieved better correction of pelvic obliquity than S hooks. Complications were almost twice as common with S hooks than screws, though this did not reach statistical significance.